A comparison of reward systems for truck drivers based

T. Bousonville I C. Ebert I J. Rath
A comparison of reward systems for truck drivers based
on telematics data and driving behavior assessments
Schriftenreihe Logistik der Fakultät für Wirtschaftswissenschaften
der htw saar
Technical reports on Logistics of the Saarland Business School
Nr. 8 (2015)
© 2015 by Hochschule für Technik und Wirtschaft des Saarlandes, Fakultät für Wirtschaftswissenschaften,
Saarland Business School
ISSN 2193-7761
A comparison of reward systems for truck drivers based on telematics data and driving behavior assessments
T. Bousonville I C. Ebert I J. Rath
Bericht/Technical Report 8 (2015)
Verantwortlich für den Inhalt der Beiträge sind die jeweils genannten Autoren.
Alle Rechte vorbehalten. Ohne ausdrückliche schriftliche Genehmigung des Herausgebers darf der Bericht oder
Teile davon nicht in irgendeiner Form – durch Fotokopie, Mikrofilm oder andere Verfahren - reproduziert werden.
Die Rechte der öffentlichen Wiedergabe durch Vortrag oder ähnliche Wege bleiben ebenfalls vorbehalten.
Die Veröffentlichungen in der Berichtsreihe der Fakultät für Wirtschaftswissenschaften können bezogen werden
über:
Hochschule für Technik und Wirtschaft des Saarlandes
Fakultät für Wirtschaftswissenschaften
Campus Rotenbühl
Waldhausweg 14
D-66123 Saarbrücken
Telefon:
E-Mail:
Internet:
+49 (0) 681/ 5867 - 519
fakultaet-wiwi@htwsaar.de
www.htwsaar.de/wiwi
A comparison of reward systems for truck drivers based on
telematics data and driving behavior assessments
Bousonville, Thomas * / Ebert, Christian ** / Rath, Jasmin *
* Hochschule für Technik und Wirtschaft des Saarlandes – University of Applied Sciences, Institut für Supply Chain
and Operations Management, Waldhausweg 14, 66123 Saarbrücken, Germany, E-mail: thomas.bousonville@htwsaarland.de
** Qivalon GmbH, Altenkesseler
christian.ebert@qivalon.de
Straße
17,
Gebäude
D2,
66117
Saarbrücken,
Germany,
E-mail:
Abstract: This paper investigates the impact of different driving behavior evaluation systems on the distribution of monetary rewards for economic driving. It starts with an introduction into truck telematics systems as the most prominent data source for the assessment
metrics. Two specific systems are presented in more detail, focusing on the way driver evaluation is performed in each of them. Data that has been collected from the two systems in a
real case is used to analyze if the drivers had a fair chance to get the same bonus independently of the system that was built in their truck.
Key words: Telematics, driving behavior, incentives, award systems
1
1
Introduction
Using data analytics on big amounts of data in order to improve the competitive position of
many businesses has become major trend. The logistics and transportation sectors do not
represent exceptions in this shift to the digital economy. Exploiting big data techniques can
help to improve operational efficiency, to provide better customer experience or to design
completely new business models (DHL 2013).
While the scientific literature on how big data will impact the logistics sector is still scarce
(Waller and Fawcett 2013), a number of applications have already made their way into daily
business and private life. A prevalent example is the navigation feature in Google Maps,
which exploits the tracking of the location data sent by millions of mobile phones in order to
determine the fluidity of traffic flows in large parts of the world.
In this paper we present the case of a trucking company that has equipped its fleet with
telematics systems that report various technical and non-technical data about the vehicle and
the driver activities in near real time to the central office. The company uses this data among
others to assess a driver’s performance regarding their more or less economic handling of
the vehicle. The goal is to award an economic driving style by an additional premium thus
setting an incentive for a win-win situation for the driver and the company.
To be able to do so the company relies on assessment metrics that are built in the telematics
systems. In practice, many fleets are composed by trucks of different vendors equipped usually with different telematics technologies. As there is no standard way to assess the economic driving behavior, the used metrics are also differing. A central question that this contribution aims to answer is if there is a fair treatment of drivers independently of the truck model
they are assigned to.
The structure of the paper is as follows. Section 2 gives an overview on modern telematics
systems and their components. In the sequel, Section 3 details the parameters that are used
to establish driving behavior assessments and explains the implemented metrics for two
widespread systems. Data from a real world case is presented and the question about the
fairness of the granted primes is discussed. The correlation of economic driving and external
operational factors are highlighted in Section 4. Section 5 summarizes and concludes the
paper.
2
Telematics systems
The term “telematics” has been coined by the French authors Simon Nora and Alain Minc for
settings emerging from the integration of computer science (“informatique”) and telecommunication technology (Nora and Minc 1978). Telematics systems comprise data storage and
processing machines as well as (usually partly wireless) network technology to exchange
2
information between them. Applications can be found in diverse industries (e.g. medicine,
education, traffic), but we focus on the use for road freight transportation, more specifically to
link the mobile vehicle via a telematics system to a remote user in the central office.
Figure 1: Possible components of a telematics system in road freight transportation
A telematics system for trucks with typical components is shown in Figure 1. The central processing unit (black box or On Board Unit - OBU) is able to determine the location of the vehicle using a satellite positioning system (usually the American Global Positioning System GPS). It is linked to other components. Via the industry standard interface FMS (Fleet Management System 2015) it gets technical data about the vehicle (like speed, position of the
accelerator pedal, total fuel consumption and many others). The digital tachograph records
the activities of the driver (“driving”, “break or rest”, “availability”, “other work”) and can be
linked to the black box through a so called D8 interface. Finally the telematics unit itself can
be equipped with a display that allows for interaction like the exchange of text messages with
the central office or the use of navigation software.
The collected data is then transferred using mobile network communication to a server that
provides the information to the end user devices, which can be dedicated client software, a
web browser or a cell phone app. The information can also be integrated via web services
into existing software solutions like transport management systems (TMS). Based on a survey on the German market, a classification of telematics systems depending on their scope
of functionalities and degree of integration has been proposed by Dudek (2013). The class
with the smallest range of functionalities provide mere localization and tracking features. On
the other side, advanced systems allow the business process oriented integration of data
captured by components illustrated in Figure 2 and potentially other information like identification data from connected hand scanners or the cooling temperature in the trailer.
The group of functionalities concerning the driver includes the management of its (remaining)
driving time (for scheduling and fleet management purposes) as well as the assessment of
3
its driving style. The latter and the way how it is used for setting up incentives for economic
driving will be discussed in the next section.
3
Assessment of driving behavior and incentive systems
There are a number of relevant parameters for the assessment of economic driving:
fuel consumption per distance, break usage, uniform speed profile, accelerator pedal
movement, number of stops and others. How these parameters are used to define a
one-dimensional measure for economic driving is depending on the telematics vendor and usually is not transparent. Even the metric of the measure can be quite different between products.
In this study we illustrate this by the comparison of two widely used systems: Daimler FleetBoard and MAN TeleMatics. The grading of economic driving behavior in FleetBoard can
vary between 1.0 and 10, the higher the better (Daimler FleetBoard 2015). The MAN solution
however assesses economic driving in per cent, 100% being the optimum (MAN 2010).
The company which provided the data for the following analysis operates a fleet with mainly
FleetBoard equipped Daimler trucks but also a significant number of MAN trucks. They were
confronted with the design problem of an incentive system that was applicable to both types
of trucks and their respective telematics systems. In general an incentive system has to
comply with the following requirements: well defined input parameters, easy understanding of
the relation between the obtained grade and the behavior of the concerned employees, the
possibility to influence the grade by adapting one’s own behavior, avoidance of unfair assessments (Schettgen 1996). An assessment would be perceived as unfair if the same performance could result in differing grades.
Table 1: Bonus categories for FleetBoard equipped vehicles
Economic
driving measure
9
9
9,3
Bonus
[€/month]
Nb of
bonuses
Nb of bonuses in %
of total
18,85%
0
161
30
124
14,52%
9,3
9,5
60
235
27,52%
9,5
9,6
90
177
20,73%
9,6
10
120
157
18,38%
Total
854
4
Table 2: Bonus categories for MAN TeleMatics equipped vehicles
Economic
driving measure [%]
Bonus
[€/month]
Nb of bonuses in %
of total
Nb of
bonuses
80%
0
41
45,56%
80%
82%
30
9
10,00%
82%
83%
60
7
7,78%
83%
84%
90
11
12,22%
100%
120
22
24,44%
Total
90
84%
As the two driving behavior metrics cannot be transformed from one into the other, the company decided to apply two different award tables, one for each group of vehicles (Table 1
and Table 2).
The columns “Economic driving measure” and “Bonus [€/month]” define the financial incentive paid based on the driver assessment from the telematics system. The remaining two
columns contain the absolute and relative number of bonuses that have been granted to
drivers in the period between January 2014 and June 2014. As can be seen from the tables,
there are 854 assessments for FleetBoard and 90 for the MAN system. The mean bonus
paid for drivers using the FleetBoard telematics was 61.58€ whereas the drivers on MAN
trucks were paid in average 48€. But not only the average premium differed significantly between the two groups, the chance to get a bonus was much higher for FleetBoard drivers (>
80%) than for the MAN driver group, where merely half of the drivers were awarded by an
extra pay. The quite different distribution of the obtained bonuses is also revealed in Figure
2. Whereas FleetBoard driver bonuses peaked at the mean bonus of 60€, MAN bonuses had
their maximum number of occurrences at the minimum and maximum values respectively.
235
Occurences
200
177
161
150
50
157
124
100
30
10
0
0
30
60
90
120
22
20
50
0
41
40
Occurences
250
0
Monthly bonus [€]
9
7
30
60
11
90
120
Monthly bonus [€]
Figure 2: Distribution of the bonuses in a) FleetBoard (left) and b) MAN TeleMatics (right)
5
The conclusion from this analysis is that the system in place during the first half of the year
2014 did not meet the requirements of objectivity and fairness. One way to cure this would
be to adapt the thresholds for getting the next premium level in one of the two systems in
order to achieve comparable means and variances of the bonuses.
4
The impact of operational difficulty
In this section another source for non-objective treatment of different drivers is investigated.
It is evident, that a truck running mainly on motorways allowing it to maintain a constant
speed with optimal rpm (motor revolutions per minute) will consume less than a vehicle
forced to many stops and accelerations. The topography and the weight of the load among
others also play an important role. Therefore a measure evaluating driver behavior claiming
some sort of objectivity has to assure that it is not (too much) biased by these factors.
The external factors that cannot be influenced by the driver but do have an impact on the fuel
consumption and tyre usage of the vehicle can be paraphrased as “operational difficulty”. A
serious provider of operations evaluations has to take this into account. This is the case for
the two systems investigated in this study. Again the scale for “operational difficulty” reaches
from 1 to 10 in the FleetBoard system (1 being the hardest) and from 0 to 100 per cent in the
Occurences
MAN TeleMatics system (100% corresponding to the highest operational demands).
300
200
100
0
FleetBoard: Dregree of operation difficulty
Occurences
60
40
20
0
MAN TeleMatics: Operation difficulty in %
Figure 3: Measurement of “operational difficulty” in a) FleetBoard (above) and b) MAN TeleMatics (below)
6
Figure 3 shows that the values for “operational difficulty” follow similar distributions (keep in
mind that for FleetBoard a small number means high operational demand, the opposite is the
case for MAN). In an unbiased system the evaluation of the driver’s performance regarding
economic driving should be independent from the “operational difficulty”. To test this hypothesis a regression analysis between the “economic driving measure” and the “operational difficulty” has been carried out for the 854 FleetBoard and 90 MAN TeleMatics data points. The
result is displayed in Figure 4.
The slope of the line should be 0 in order to reflect independency of the driving assessment
from external factors (“operational difficulty”). This is obviously not the case for either one of
the two systems. In Figure 4 a) medium or bad grades for economic driving are not frequent
but possible in combination with low operational difficulty (smaller than 5.5), while higher operational difficulty (above of 5.5) systematically leads to lower driving evaluations. A similar
observation can be made for MAN TeleMatics, where an increase of 1% in “operational diffi-
Economic driving (1‐10)
culty” statistically leads to a 0.5% decrease in the assessment of economic driving.
10
8
6
y = ‐0.4645x + 11.179
R² = 0.3009
4
2
0
0
2
4
6
8
10
80
100
Economic driving (%)
Operational difficulty (1‐10)
100
80
60
y = 0.4875x + 44.52
R² = 0.5692
40
20
0
0
20
40
60
Operational difficulty (%)
Figure 4: Regression analysis on „Economic driving measure“ against „Operational difficulty“ for a)
FleetBoard data (above) and b) MAN TeleMatics data (below)
7
5
Conclusion
In this contribution gathering data by telematics systems and using it subsequently in human
resource management is presented as an example for a “big data” application in transportation. Main technological components and modules of telematics systems are introduced with
a focus on elements dealing with driving style assessment. For two concrete systems the
metrics for driving assessment are presented. The incentive systems that have been implemented by the company of the case study were compared regarding objectivity, i.e. the fair
chance for a driver to get a bonus independently from the vehicle he is assigned to.
The findings suggest that this was not the case for the investigated period. However, directions how to improve the equity of the incentive systems can be derived from the analysis.
Thus, a major conclusion is that companies applying incentive systems that are built on top
of heterogeneous telematics assessments of driving performance should check their reward
systems based on statistical analysis like the one presented on this paper. Another source of
bias is the impact of external factors on the grading of economic driving. Using linear regression it had been shown that the possible grade for “economic driving” is depending on the
operational environment, especially as this is getting more demanding.
Acknowledgments
The authors thank the participating trucking company for providing the data that made the
investigation possible.
REFERENCES
DHL (2013) Big Data in Logistics, DHL Customer Solutions and Innovations, Troisdorf, 2013.
Daimler FleetBoard (2015) User Manual FleetBoard Cockpit, Release: 01/2015 1.17.0aEN, Stuttgart, 2015.
Dudek (2013) Telematik 2012, Duale Hochschule Baden Württemberg, Ravensburg, 2013.
Fleet Management System (2015) FMS Standard Homepage, www.fms-standard.com, Accessed on March 01, 2015.
MAN (2010) MAN TeleMatics Benutzerhandbuch, Version 04.06.2010, 2010.
Nora and Minc (1978) L'informatisation de la société, Paris, 1978.
8
Schettgen (1996) Arbeit, Leistung, Lohn. Analyse und Bewertungsmethoden aus sozioökonomischer Perspektive, Stuttgart, 1996.
Waller and Fawcett (2013) Data Science, Predictive Analytics, and Big Data: A Revolution
That Will Transform Supply Chain Design and Management, Journal of Business Logistics
34, 77-84.
9
Veröffentlichte Berichte der Fakultät für Wirt-
Published reports of the Saarland Business
schaftswissenschaften
School
Die PDF-Dateien der folgenden Berichte sind verfüg-
The PDF files of the following reports are available
bar unter:
under:
http://www.htwsaar.de/wiwi
1
I. Correia, T. Melo, F. Saldanha
da Gama
Comparing classical performance
Logistik-Technologien,
KMU
Mittelstand,
Logistik-Technologien,
KMU
(5 pages, 2012)
(6 pages, 2014)
4
7
Mittelstand,
measures for a multi-period, twoechelon
supply
chain
network
A. Steinhaus, S. Hütter
design problem with sizing deciLeitfaden zur Implementierung von
RFID in kleinen und mittelständischen Unternehmen
sions
Keywords: supply chain network
design, facility location, capacity
acquisition, profit maximization, cost
minimization
Keywords: RFID, KMU, schlanke
Prozesse, Prozessoptimierung, Produktion, Forschungsgruppe Qbing
(43 pages, 2012)
(49 pages, 2013)
2
5
T. Melo
S.A.
Alumur,
M.T. Melo
B.Y.
Kara,
tics network over multiple time
selection and outsourcing opportunities
Keywords: logistics network design/re-design,
multiple
periods,
transportation mode selection, product outsourcing, mixed-integer linear
programming
8
Forschung an Fachhochschulen –
im
Mit-
telstand: Ergebnisse der QbingTrendumfrage 2013
Keywords:
Innovation,
Umfrage,
Trendbarometer, Logistik-Konzepte,
Forschung an Fachhochschulen –
für
Innovation
truck
drivers
based
on
telematics data and driving behav-
S. Hütter, A. Steinhaus
Treiber
T. Bousonville, C. Ebert, J. Rath
A comparison of reward systems
for
6
S. Hütter, A. Steinhaus
Innovation
Keywords: forward logistics network
design, reverse logistics network
design, models, applications
(26 pages, 2013)
(13 pages, 2012)
für
Redesigning a three-echelon logis-
(49 pages, 2014)
Keywords: Hospital logistics, Operations Research, application areas
Treiber
Lopes,
Location and Logistics
hospital logistics
3
M.J.
periods with transportation mode
A note on challenges and opportunities for Operations Research in
M.J. Cortinhal,
M.T. Melo
im
Mit-
telstand: Ergebnisse der Qbing-
ior assessments
Keywords: Telematics, driving behavior, incentives, award systems
Trendumfrage 2014
(9 pages, 2015)
Keywords:
Innovation,
Umfrage,
Trendbarometer, Logistik-Konzepte,
Hochschule für Technik und Wirtschaft des Saarlandes
Die Hochschule für Technik und Wirtschaft des Saarlandes (htw saar) wurde im Jahre 1971 als saarländische
Fachhochschule gegründet. Insgesamt studieren rund 5000 Studentinnen und Studenten in 38 verschiedenen
Studiengängen an der htw saar, aufgeteilt auf vier Fakultäten.
In den vergangenen zwanzig Jahren hat die Logistik immens an Bedeutung gewonnen. Die htw saar hat dieser Entwicklung frühzeitig Rechnung getragen und einschlägige Studienprogramme sowie signifikante Forschungs- und Technologietransferaktivitäten entwickelt. Die Veröffentlichung der Schriftenreihe Logistik soll
die Ergebnisse aus Forschung und Projektpraxis der Öffentlichkeit zugänglich machen.
Weitere Informationen finden Sie unter http://logistik.htwsaar.de
Institut für Supply Chain und Operations Management
Das Institut für Supply Chain und Operations Management (ISCOM) der htw saar ist auf die Anwendung quantitativer Methoden in der Logistik und deren Implementierung in IT-Systemen spezialisiert. Neben öffentlich
geförderten Forschungsprojekten zu innovativen Themen arbeitet ISCOM eng mit Projektpartnern aus der
Wirtschaft zusammen, wodurch der Wissens- und Technologietransfer in die Praxis gewährleistet wird. Zu den
Arbeitsgebieten zählen unter anderem Distributions- und Transportplanung, Supply Chain Design, Bestandsmanagement in Supply Chains, Materialflussanalyse und -gestaltung sowie Revenue Management.
Weitere Informationen finden Sie unter http://iscom.htwsaar.de
Forschungsgruppe Qbing
Qbing ist eine Forschungsgruppe an der Hochschule für Technik und Wirtschaft des Saarlandes, die spezialisiert ist auf interdisziplinäre Projekte in den Bereichen Produktion, Logistik und Technologie. Ein Team aus
derzeit acht Ingenieuren und Logistikexperten arbeitet unter der wissenschaftlichen Leitung von Prof. Dr. Steffen Hütter sowohl in öffentlich geförderten Projekten als auch zusammen mit Industriepartnern an aktuellen
Fragestellungen zur Optimierung von logistischen Prozessabläufen in Handel und Industrie unter Einbeziehung modernster Sensortechnologie und Telemetrie. Qbing hat auch und gerade auf dem Gebiet der angewandten Forschung Erfahrung in der Zusammenarbeit mit kleinen und mittelständischen Unternehmen.
Weitere Informationen finden Sie unter http://www.qbing.de
ISSN 2193-7761